[p2p-research] Is the "lump of labor fallacy" itself a fallacy?

Paul D. Fernhout pdfernhout at kurtz-fernhout.com
Sat Nov 28 20:38:15 CET 2009

Ryan Lanham wrote:
> On Sat, Nov 28, 2009 at 11:02 AM, Paul D. Fernhout <
> pdfernhout at kurtz-fernhout.com> wrote:
>> But, the difference is presumably the claim that industrial productivity
>> has increased. I don't know that figure. Here it says 2.1% increanes in
>> output per person hour from 1937 to 1952:
>>  http://www.jstor.org/pss/2550357
>> Taking that from 1938 to now (just assuming it as a constant for 70 years,
>> compounded) that is about a factor of four. So, that would suggest that in
>> today's dollars, a minimum wage should be 4 * US$3.64 or US$14.56.
> Sadly, in industrialized societies, especially post-industrial societies,
> most productivity gain occurs at the top...not at the bottom of wages.  Part
> of the problem with capitalism is that most people are nearly worthless in
> capital production terms.  Skills increase scarcity dramatically and are
> typically channeled into high profit/high wage functions.  Rare is simply
> better in most cases in market systems.
> That is of course what makes abundance theory so interesting.
> Highly skilled and esoteric workers are not hurting for wages or work.  The
> people hurting for work are those with modest skills and who do things most
> people can do.

These are really good points. And they fit with the notion of "technology as 
an amplifier" which I first wrote about in college in the 1980s (inspired in 
part from reading about cybernetics). So, technology is able to amplify the 
high skilled worker (whatever high skill means at the time), and so there is 
less need to amplify the low skill worker. So, there is almost a "winner 
takes all" on wages.

Or there would be, except for at least two other issues.

If only a few high skilled workers get a lot, unions won't so easily form 
for those jobs. There is not "programmers union" for example, though there 
are unions that may include some jobs descriptions that entail some 
programming). Without a union, it is difficult for these workers at the top 
to get a proportionate share of what they produce. They may get a lot, and a 
lot more than most other workers, but it still may be a tiny fraction of the 
value they really add.

Also, the fact that there are many workers in general puts a sociological 
downward pressure on high end wages. For example, even if there can be 
orders of magnitude difference in programmer productivity (including some 
programmers being a net negative to a project, see the Mythical Man Month), 
the programmers who may be literally 1000X as productive as some others 
usually get maybe at most 2X or 3X more salary. And if companies paid the 
star programmers more, it would breed social resentment of all sorts, and 
there would also be a great pressure at the bottom to raise wages. Some very 
few companies (Microsoft or Google) are places where programmers or anyone 
could make a lot of money also through stock options, but stock 
participation rarely has much to do with productivity and more to do with 
luck and timing.

Some of this is also true because it is very difficult to assess programmer 
productivity sometimes (or that of almost any other creative endeavor). How 
can you set salaries related to performance when most people have no way of 
judging the quality of output of a programmer. For example, the programmer 
who says, "Let's use the XYZ library" may be literally 1000X more productive 
than the one who writes 10000 lines of buggy and poorly tested code, but the 
one who wrote the most code will probably be rewarded more as being more 

The movie "The Seven Samurai" shows an example of this problem -- the 
peasants want to hire Samurai to defend their village, but they have no idea 
what makes for a good Samurai. Swashbuckling sword play? A minimum of 
effort? There is a scene where the peasants watch two Samurai duel, for 
example, arguing over some point of technique -- although in that case the 
winner was obvious, since only one lived. Fortunately, programming is not 
(usually) so deadly a profession -- although vitamin D deficiency and 
sedentary work in a chair can probably make it so for many programmers, as 
can the results of programming cause deaths for others through military 
robotics, malfunctioning aircraft and medical devices, and even bureaucratic 
processes that keep people from health care, and so on. Sadly, the most 
money I've ever been paid for developing software as a consultant was 
fifteen years ago for a project to help deny people health care (via 
supporting more accurate telephone interviews for insurance applications). 
Is that really productive compared to better medical records to support 
better health care advice? So, one needs to factor in rewarding people for 
creating artificial scarcity too when one analyzes what a productive 
programmer really is. I'm sure the same would apply for other professions as 

--Paul Fernhout

More information about the p2presearch mailing list